annotate tools/rdock/bin/sdreport @ 2:0faa03a92843 draft default tip

Uploaded

#!/usr/bin/perl
# Produces text summaries of SD records                                        *
#                                                                              *
# Usage:  sdreport [-l] [-t] [-s<DataField>] [sdFiles]                         *
#                                                                              *
#         -l output data fields for each record as processed                   *
#         -t tabulate Rbt.Score.* fields for each record as processed          *
#         -s summarise data fields for all records                             *
#         -s<DataField> summarise data fields for each unique value            *
#                       of <DataField>                                         *
#                                                                              *
# Note:   If -l or -t are combined with -s, the listing/table is output        *
#         within each ligand summary                                           *
#         If sdFiles not given, reads from standard input                      *
#         Output is to standard output                                         *
#         Default is equivalent to sdfilter -l                                 *
#*******************************************************************************
use lib "$ENV{'RBT_ROOT'}/lib";

use SDRecord;

# Default field names and headings for rDock v4.00 scores
my @defaultFields = ('SCORE','SCORE.INTER','SCORE.INTRA','SCORE.RESTR','SCORE.INTER.VDW');
my @defaultHeadings = ('TOTAL','INTER','INTRA','RESTR','VDW');

# Default field names and headings for normalised scores (score / #ligand heavy atoms)
my @defaultNormFields = ('SCORE.norm','SCORE.INTER.norm','SCORE.INTRA.norm','SCORE.RESTR.norm','SCORE.heavy');
my @defaultNormHeadings = ('TOTALn','INTERn','INTRAn','RESTRn','#heavy');

# Default field names and headings for rDock v3.00 scores
my @defaultOldFields = ('Rbt.Score.Corrected','Rbt.Score.Inter','Rbt.Score.Intra','Rbt.Score.IntraMin','Rbt.Score.Restraint');
my @defaultOldHeadings = ('TOTAL','INTER','INTRA','INTRAMIN','RESTR');

my $listFormat = 0;
my $summaryFormat = 0;
my $tableFormat = 0;
my $supplierFormat = 0;
my $csvFormat = 0;
my $summaryKey = "_TITLE1";
my $oldFields = 0;#If true, use old default field names for component scores
my $headings = 1;#DM 21 Nov 2000, If false, don't output headings
my @outputFields;
my @outputHeadings;

#Print help if no command line arguments
printHelpAndExit() if (scalar(@ARGV) == 0);

#Parse command line arguments
my $nArgs = scalar(@ARGV);

while (scalar(@ARGV)) {
  my $arg = shift @ARGV;
  printHelpAndExit() if ($arg eq '-h');
  if (index($arg,'-l')==0) {
    $listFormat = 1;
  }
  elsif (index($arg,'-o')==0) {
    $oldFields = 1;
  }
  # 7 Feb 2005 (DM) Option to report normalised aggregate scores
  elsif (index($arg,'-norm')==0) {
    $oldFields = 2;
  }
  elsif (index($arg,'-sup')==0) {
    $supplierFormat = 1;
  }
  elsif (index($arg,'-s')==0) {
    $summaryFormat = 1;
  }
  elsif (index($arg,'-id')==0) {
    $summaryKey = substr($arg,3);
  }
  elsif (index($arg,'-nh')==0) {
    $headings = 0;
  }
  elsif (index($arg,'-t')==0) {
    $tableFormat = 1;
    push @outputFields, split(',',substr($arg,2));
    push @outputHeadings, @outputFields;
  }
  elsif (index($arg,'-c')==0) {
    $csvFormat = 1;
    push @outputFields, split(',',substr($arg,2));
    push @outputHeadings, @outputFields;
  }
  else {
    push @files,$arg;#must be a filename
  }
}
push @ARGV,@files;#put the filenames back in the arg list

#use -l if neither table format is specified
$listFormat = (!$tableFormat && !$csvFormat && !$supplierFormat);

#If no output fields defined for -t or -c use the defaults (old or new)
if (scalar(@outputFields)==0) {
	if ($oldFields == 1) {
		@outputFields = @defaultOldFields;
		@outputHeadings = @defaultOldHeadings;
	}
	elsif ($oldFields == 2) {
                @outputFields = @defaultNormFields;
                @outputHeadings = @defaultNormHeadings;
	}
	else {
		@outputFields = @defaultFields;
		@outputHeadings = @defaultHeadings;
	}
}

my $sdRec = new SDRecord;
my %summary;#hash of SDRecord lists, indexed by user-defined summary key
my %indexByName;
my %indexByNum;
my $idx = 0;
my $nRec = 0;

#Column headings for tab and csv format
#DM 21 Nov 2000 - if $headings is false, then don't output the column headings
if ($tableFormat && !$summaryFormat && $headings) {
  tabHeadings($summaryKey,@outputHeadings);
}
if ($csvFormat && !$summaryFormat && $headings) {
  csvHeadings($summaryKey,@outputHeadings);
}

#read records
while ($sdRec->readRec('LINES'=>1,'DATA'=>1)) {
  $sdRec->addData('_REC' => ++$nRec);#add record# as temp data field
  if ($listFormat && !$summaryFormat) {
    print "\n\nRECORD #$nRec\n";
    $sdRec->writeData();
  }
  if ($tableFormat && !$summaryFormat) {
    @recList = ($sdRec);
    tabScores(\@recList,$summaryKey,@outputFields);
  }
  elsif ($csvFormat && !$summaryFormat) {
    @recList = ($sdRec);
    csvScores(\@recList,$summaryKey,@outputFields);
  }
  elsif ($supplierFormat && !$summaryFormat) {
    @recList = ($sdRec);
    tabulateSuppliers(\@recList,$summaryKey);
  }
  #add record to summary, indexed by user field value
  #keep a separate index of unique values of user field values,
  #indexed by number in the order the values first appear
  #In this way we can maintain the sorted order of ligands
  #when we come to print out the summary
  if ($summaryFormat) {
    my $summaryValue = $sdRec->{'DATA'}->{$summaryKey};
    #New data field value encountered
    if (!defined $indexByName{$summaryValue}) {
      $idx++;#incr the number of unique values
      #keep cross-referenced indexes (field value <-> number)
      $indexByName{$summaryValue} = $idx;
      $indexByNum{$idx} = $summaryValue;
    }
    push @{$summary{$summaryValue}},$sdRec->copy('DATA'=>1);
  }
}

#Print summary if required
if ($summaryFormat) {
  print "\n===============================================================\n";
  print "SUMMARY BY $summaryKey\n";
  foreach $idx (sort {$a<=>$b} keys %indexByNum) {#numberic sort of index numbers
    my $key = $indexByNum{$idx};#look up corresponding data field value
    print "\n===============================================================\n";
    print "$summaryKey = $key (#$idx)\n\n";
    writeSummary($summary{$key});
    if ($listFormat) {
      print "\nIndividual records:\n";
      foreach $rec (@{$summary{$key}}) {
        print "\n";
	      $rec->writeData();
      }
    }
    if ($tableFormat) {
      print "\nScores:\n";
      tabHeadings($summaryKey,@outputHeadings);
      tabScores($summary{$key},$summaryKey,@outputFields);
    }
    if ($csvFormat) {
      print "\nScores:\n";
      csvHeadings($summaryKey,@outputHeadings);
      csvScores($summary{$key},$summaryKey,@outputFields);
    }
  }
}

##############################################################
# writes a summary to STDOUT for a list of SDRecords
# Input is a reference to an array of SDRecords
sub writeSummary {
  my $recListRef = shift;

  #Extract the list of data values for each fieldname into a hash array
  #(key=fieldname, value=array ref)
  my %fields;
  foreach $rec (@{$recListRef}) {
    my ($key,$value);
    while ( ($key,$value) = each %{$rec->{'DATA'}}) {
      push @{$fields{$key}},$value;
    }
  }

  #Look for constant fields and store separately
  my %constFields;
  foreach $key (keys %fields) {
    my @values = sort  @{$fields{$key}};
    my $nVal = scalar(@values);
    if ($values[0] eq $values[$nVal -1]) {
      $constFields{$key} = $values[0];#store the field name and the constant value
      delete $fields{$key};#remove from (non-const) array
    }
  }

  #Print constant fields
  print "\nConstant fields:\n\n";
  foreach $key (sort keys %constFields) {
    printf "%-40s%s\n",$key,$constFields{$key};
  }
  #Print min and max value for non-const fields
  print "\nVariable fields:\n\n";
  foreach $key (sort keys %fields) {
    my @values = @{$fields{$key}};
    #Look at first value to decide whether to do text or numeric sort
    if (isNaN($values[0])) {
      @values = sort @values;#text sort
    }
    else {
      @values = sort {$a <=> $b} @values;#numeric sort
    }
    my $nVal = scalar(@values);
    printf "%-40s",$key;
    print "Min = $values[0]\tMax = $values[$nVal-1]\t(N = $nVal)\n";
  }
}

##############################################################
# function isNaN equivalent to the C++, java, javascript isNaN
# From P Vaglio, ~intranet/lib/rbt_func.pl
sub isNaN {
  local($_) = @_;
  s/\s+//g;         # strip white space
  # match +or- beginning of line 0 or 1 time
  # then any numeric 0 or more
  # then a decimal point 
  # then any numeric 0 or more after decimal point 
  # then possibly an e or E then + or - then any numreci at least once
  if (/^([+-]?)(\d*)(\.(\d*))?([Ee]([+-]?\d+))?$/ && (defined $2 || defined $4)) {
    return 0;
  } else {
    return 1;
  }
}

##############################################################
# output corresponding headings for use with tabScores
sub tabHeadings {
  my $summaryKey = shift;
  my @fieldNames = @_;
  printf("%-10s%-30s","REC",$summaryKey);
  foreach $field (@fieldNames) {
    printf("%10s",$field);
  }
  print "\n";
}

##############################################################
# tab-delimited output of named data field values
sub tabScores {
  my $recListRef = shift;
  my $summaryKey = shift;
  my @fieldNames = @_;
  foreach $rec (@{$recListRef}) {
    printf("%03d\t%-30.30s",$rec->{'DATA'}->{'_REC'},$rec->{'DATA'}->{$summaryKey});
    foreach $field (@fieldNames) {
      my $val = $rec->{'DATA'}->{$field};
      if (isNaN($val)) {
        printf("%-10.12s",$val);
      }
      else {
        printf("%10.3f",$val);
      }
    }
    print "\n";
  }
}

##############################################################
# output corresponding headings for use with csvScores
sub csvHeadings {
  my $summaryKey = shift;
  my @fieldNames = @_;
  printf("%s,%s","REC",$summaryKey);
  foreach $field (@fieldNames) {
    printf(",%s",$field);
  }
  print "\n";
}

##############################################################
# comma-delimited output of named data field values
sub csvScores {
  my $recListRef = shift;
  my $summaryKey = shift;
  my @fieldNames = @_;
  foreach $rec (@{$recListRef}) {
    printf("%d,%s",$rec->{'DATA'}->{'_REC'},$rec->{'DATA'}->{$summaryKey});
    foreach $field (@fieldNames) {
      my $val = $rec->{'DATA'}->{$field};
      if (isNaN($val)) {
        printf(",%s",$val);
      }
      else {
        printf(",%.3f",$val);
      }
    }
    print "\n";
  }
}


##############################################################
# standardised output of Catalyst supplier field
# Input is a reference to an array of SDRecords
# and a ligand identifier field to output in column 1 (def=Name)
sub tabulateSuppliers {
  my $recListRef = shift;
  my $summaryKey = shift || 'Name';
  foreach $rec (@{$recListRef}) {
    my $suppBase = $rec->{'DATAREF'}->{'Supplier'}+1;
    my $linesRef = $rec->{'LINES'};
    my $name = $rec->{'DATA'}->{$summaryKey};

    #Output some useful compound info
    my $name = $rec->{'DATA'}->{$summaryKey};
    my $molFormula = $rec->{'DATA'}->{'MolFormula'};
    my $molWt = $rec->{'DATA'}->{'MolWt'};
    my $casNum = $rec->{'DATA'}->{'CAS_num'};
    my $mdlNum = $rec->{'DATA'}->{'MDLNUMBER'};
    print "\n\n====================================================================================================\n";
    printf("%-10.10s%s\n","Name:",$name);
    printf("%-10.10s%s\n","Formula:",$molFormula);
    printf("%-10.10s%s\n","Mol.wt:",$molWt);
    printf("%-10.10s%s\n","CAS #:",$casNum);
    printf("%-10.10s%s\n","MDL #:",$mdlNum);

    #Get all the supplier record lines into a list
    #Record is terminated by blank line
    my @lines;
    my $nLines = 0;
    for (; $$linesRef[$suppBase+$nLines] ne ""; $nLines++) {
      push @lines,$$linesRef[$suppBase+$nLines];
    }

    #Column headings
    printf("\n%-20.20s%-40.40s%-40.40s\n",
	   "Supplier",
	   "Comment",
	   "Price"
	  );
    print "----------------------------------------------------------------------------------------------------\n";

    #Loop over each supplier
    my $iLine = 0;
    for (; $iLine < $nLines; $iLine++) {
      #collect supplier info lines
      my @supplierInfo = ();
      for (; $lines[$iLine] ne "." && $iLine < $nLines; $iLine++) {
	push @supplierInfo,$lines[$iLine];
      }
      #Check for incomplete record
      if ($iLine == $nLines) {
	print "** INCOMPLETE RECORD **\n";
	last;
      }
      my $nSupplierInfo = scalar(@supplierInfo);
      my $supplier = $supplierInfo[0];
      #loop over each grade
      for ($iLine++; ($lines[$iLine] ne "........................") && ($iLine < $nLines); $iLine++) {
	#collect grade info lines
	my @gradeInfo = ();
	for (; index($lines[$iLine],"_") ne 0 && $iLine < $nLines; $iLine++) {
	  push @gradeInfo,$lines[$iLine];
	}
	#Check for incomplete record
	if ($iLine == $nLines) {
	  print "** INCOMPLETE RECORD **\n";
	  last;
	}
	my $grade = $gradeInfo[0];
	#loop over each price info line
	for (; index($lines[$iLine],".") ne 0 && $iLine < $nLines; $iLine++) {
	  my @priceInfo = split(" ",$lines[$iLine]);
	  my $price = join(" ",@priceInfo);
	  printf("%-20.20s%-40.39s%-40.40s\n",
		 $supplier,
		 $grade,
		 $price);		 
	}
	#Check for incomplete record
	if ($iLine == $nLines) {
	  print "** INCOMPLETE RECORD **\n";
	  last;
	}
	last if $lines[$iLine] eq "........................";
      }
    }
  }
}


#######################################################################
sub printHelpAndExit {
  print "\nProduces text summaries of SD records\n";
  print "\nUsage:\tsdreport [-l] [-t[<FieldName,FieldName...>]] [-c<FieldName,FieldName...>] [-id<IDField>] [-nh] [-o] [-s] [-sup] [sdFiles]\n";
  print "\n\t-l (list format) output all data fields for each record as processed\n";
  print "\t-t (tab format) tabulate selected fields for each record as processed\n";
  print "\t-c (csv format) comma delimited output of selected fields for each record as processed\n";
  print "\t-s (summary format) output summary statistics for each unique value of ligand ID\n";
  print "\t-sup (supplier format) tabulate supplier details (from Catalyst)\n";
  print "\t-id<IDField> data field to use as ligand ID\n";
  print "\t-nh don't output column headings in -t and -c formats\n";
  print "\t-o use old (v3.00) score field names as default columns in -t and -c formats, else use v4.00 field names\n";
  print "\t-norm use normalised score field names as default columns in -t and -c formats (normalised = score / #ligand heavy atoms)\n";
  print "\nNote:\tIf -l, -t or -c are combined with -s, the listing/table is output within each ligand summary\n";
  print "\t-sup should not be combined with other options\n";
  print "\tDefault field names for -t and -c are rDock score field names\n";
  print "\tDefault ID field name is Name\n";
  print "\n\tIf sdFiles not given, reads from standard input\n";
  print "\tOutput is to standard output\n\n";
  exit;
}